Jaeckle, Kurt A.

1520 Background: Multiple variables may impact endpoints utilized to assess efficacy of investigational therapies for pts with recurrent gliomas. Understanding the nature of these variables should aid interpretation of results. Methods: 583 pts were enrolled to 15 consecutive recurrent glioma clinical trials conducted between 1980 and 2004. Identical patient, tumor, treatment and time variables were identified for each trial. Central neuropathology review was completed and pts were grouped by WHO diagnoses. Pts alive and progression-free at 6 months after start of treatment were a PFS6 ‘success’. Pts whose outcome was progression at first evaluation were classified as “Immediate Progression (ImmProg)”. Response was defined as regression, partial regression, or complete regression in the absence of worsening neurologic symptoms or increased corticosteroid dose. Univariable and multivariable Classification and Regression Trees (CART) models were used to identify factors associated with OS, PFS, PFS6, ImmProg, and Response. Results: Variables associated with outcomes for each endpoint assessed by CART models are shown in Table below. Conclusions: In clinical trials of recurrent glioma, age, PS, duration of disease, primary and recurrent histologic diagnosis, and primary and recurrent extent of resection are associated with outcomes. These variables should be validated in additional datasets for potential inclusion in reports of future clinical trials. [Table: see text] No significant financial relationships to disclose.

2075 Background: Multiple variables may impact endpoints used to assess therapeutic efficacy for patients with recurrent gliomas. To identify prognostic factors for each of 3 outcome variables, a master analysis file was constructed by pooling datasets summarizing key baseline factors and outcomes obtained in 15 North Central Cancer Treatment Group (NCCTG) trials (n=583, 1980–2004) and 12 North American Brain Tumor Consortium (NABTC) trials (n=596, 1998–2002). Methods: In addition to traditional variables (gender, age, performance score (PS), race, extent of primary resection, year of study entry, last known histology, time since initial diagnosis, baseline steroid use, prior chemotherapy (Y/N), prior nitrosoureas, baseline anticonvulsant use), we investigated prior t (TMZ), number of relapses, academic vs. community site and low grade at initial diagnosis. Outcome variables were: (1) overall survival (OS), (2) progression free survival (PFS) and (3) PFS6 = proportion of patients alive and progression-free 6 months after start of treatment. For all 3 endpoints, Classification and Regression Tree (CART) models and bootstrap samples were used to identify prognostic variables and to look for evidence of interactions among the independent variables affecting outcome. Results: For all outcomes the initial CART analysis selected last known grade (4 vs. 3) as the most important variable. Additional factors selected as predicting poor prognosis differed with grade and outcome measure and included prior TMZ, longer time since primary diagnosis, older age and poor PS. Community vs. academic was not a strong predictor in any of the models nor was low grade at initial diagnosis. Additional analyses are planned to better describe the impact of all the variables on outcome, including use of Cox models and logistic models considering all baseline variables with backwards selection. Conclusions: The results from analyses to date are consistent with current thinking that last known histology (grade) is an important factor. Time since initial diagnosis is likely a surrogate for number of relapses. More definitive conclusions await the completion of ongoing analyses. No significant financial relationships to disclose.

2004 Background: Vorinostat is a histone deacetylase inhibitor that represents a rational therapeutic target in GBM. Methods: Recurrent GBM patients (pts) having received = 1 prior chemo regimen for progressive disease were eligible. Vorinostat dose was 200 mg bid x 14 days q 3 weeks. We applied a phase II design which had 90% power to declare the regimen active if the study PFS6 rate was 25%, a 15% rate increase over PFS6 in historical controls (10%). Results: 68 patients were treated. Grade 3+ non-hematologic toxicity consisting mainly of fatigue, diarrhea and dehydration occurred in 23% of pts; Grade 3+ hematologic toxicity consisting mainly of thrombocytopenia occurred in 25% of pts. Intra-patient dose escalation to 300 mg bid resulted in higher incidence of grade 3+ thrombocytopenia (45%) and was aborted. Pts receiving enzyme-inducing anticonvulsants (EIAC) had significantly less grade 3 + non-hematologic toxicity (p=0.01) and grade 3/4 thrombocytopenia (p=0.04). Pharmacokinetic analysis showed lower vorninostat C-max and AUC (0–24h) values and higher vorinostat-glucuronide C-max and AUC (0–24h) values in EIAC pts, although this did not reach statistical significance. The trial met the prospectively defined primary efficacy endpoint at the planned interim analysis with 5 of the first 22 patients (23%) being progression-free at 6 mo. RNA array analysis, performed in paired baseline and post-vorinostat treatment samples in a separate subgroup of 5 surgical recurrent GBM pts, who received vorinostat for 6 doses prior to surgery, showed upregulation of E-cadherin (p=0.02), thus indicating a biologic effect of the HDAC inhibitor on the glioblastoma tumors. Histone acetylation, Akt, phospho-Akt, P21Waf1/Cip1 and P27Kip1 expression analysis in all patients and correlation with outcome is ongoing. Conclusions: Vorinostat is well tolerated in recurrent GBM patients. EIAC patients have less toxicity, likely due to increased vorinostat metabolism via glucuronidation. Interim efficacy analysis is indicative of antitumor activity. Final efficacy analysis, RNA array data and correlative tissue analysis will be reported. No significant financial relationships to disclose.

Abstract Background Cognitive function is an important outcome in brain-tumor clinical trials. Cognitive examiners are often needed across multiple sites, many of whom have no prior testing experience. To ensure quality, we looked at examiner errors in administering a commonly used cognitive test battery, determined whether the errors were correctable upon central review, and considered whether the same errors would be detected using onsite electronic data entry. Methods We looked at 500 cognitive exams administered for brain-tumor trials led by the Alliance for Clinical Trials in Oncology (Alliance). Of 2277 tests examined, 32 noncorrectable errors were detected with routine central review (1.4% of tests administered), and thus removed from the database of the respective trial. The invalidation rate for each test was 0.8% for each part of the Hopkins Verbal Learning Test-Revised, 0.8% for Controlled Oral Word Association, 1.8% for Trail Making Test-A and 2.6% for Trail Making Test-B. It was estimated that, with onsite data entry and no central review, 4.9% of the tests entered would have uncorrected errors and 1.3% of entered tests would be frankly invalid but not removed. Conclusions Cognitive test results are useful and robust outcome measures for brain-tumor clinical trials. Error rates are extremely low, and almost all are correctable with central review of scoring, which is easy to accomplish. We caution that many errors could be missed if onsite electronic entry is utilized instead of central review, and it would be important to mitigate the risk of invalid scores being entered. ClinicalTrials.gov identifiers NCT01781468 (Alliance A221101), NCT01372774 (NCCTG N107C), NCT00731731 (NCCTG N0874), and NCT00887146 (NCCTG N0577).